An ontology-enabled cloud resource management framework incorporating user trust evaluation and dynamic virtual machine allocation for load balancing and migration control

Abstract Cloud computing supports data-intensive applications by providing efficient resource management and task scheduling. However, existing resource allocation methods often struggle to handle dynamic workloads, heterogeneous virtual machine (VM) environments, and changing user behavior. To address these challenges, this paper proposes an Ontology-Driven Trust-Based Virtual Machine Allocation and Migration with Fuzzy Logic (OTVAM-FL) framework. The proposed framework combines ontology-based semantic matching, user trust evaluation, and fuzzy logic to improve VM allocation, task scheduling, and migration decisions. First, a User Reputation Score (URS) is calculated from historical task success, completion time, and failure rate to prioritize user tasks. An ontology model is then used to represent users, tasks, and virtual machines, enabling semantic matching between task requirements and available resources. Fuzzy logic uses CPU utilization, RAM utilization, and bandwidth utilization to support resource allocation and VM migration under changing workload conditions. The proposed framework was evaluated using the CloudSim simulation environment and compared with MSGO-RNN, I-ANFIS, and SEBF-EIOT. The results show that OTVAM-FL achieves lower CPU utilization, lower RAM utilization, reduced bandwidth usage, fewer VM migrations, and shorter task execution time than the compared methods under the evaluated simulation settings. These results demonstrate that the proposed framework improves resource utilization and task scheduling efficiency in a simulated cloud environment.